1920] 
DUGGAR—H-ION CONCENTRATION AND NUTRIENT SOLUTIONS 43 
considerable diversity in the intervals between the renewal of 
solutions. These intervals have varied from four days to a time 
interval covering the entire culture period. The results have 
. been as consistent as might be expected, and it is believed that 
the value of lengthening the interval in this way is important 
from the standpoint of reducing the labor in the maintenance of 
such cultures. Trelease and Free (^17) have, however, shown 
that frequent changes of solution are more favorable for growth; 
but it also appears from their data that renewals during the 
first two weeks are not so important as those made later. They 
suggest that a continuous flow of solution through the culture 
is more beneficial than a daily change. 
The work of Pantanelli (15) and others has shown that after 
plants have been for a few hours in contact with salt solutions 
it may be demonstrated that there has been a different rate of 
absorption of the various ions. "The solution, therefore, changes 
rapidly in the presence of abundant absorbing surfaces. Hoag- 
land has also emphasized this point, and within a certain range 
of osmotic concentration he regards the initial concentration of 
any particular ion as practically immaterial. Nevertheless, it 
would be admitted by all that there must be on the one hand a 
true physiological balance, and that on the other hand, the 
concentration of no necessary ion or molecule shall become a 
limiting factor in growth. It is not, however, proposed to dis- 
cuss in this paper the significance and final results of the inter- 
change of ions or molecules between roots and solutions. 
Discussing limiting factors in water cultures Stiles (16) has 
drawn attention to the limited application of the water-culture 
method in physiological problems. He regards this as related 
to “ (1) the difficulty in analyzing results due to the complex of 
factors not under control; (2) the difficulty of controlling in 
some cases even the factor whose action is being investigated; 
and (3) the excess of labor required to produce results which are 
only of a low degree of accuracy.” 
In a previous paper Stiles (15) has in a measure erystallized 
the feeling of many investigators working with water cultures in 
arriving at the conclusion that it is necessary to calculate the 
probable error of the results in accurately evaluating the sig- 
nificance of differences exhibited by different sets of cultures. 
